Naltrexone is an opioid receptor antagonist used primarily in the management of alcohol dependence and opioid dependence. It is available on the market in form of its hydrochloric salt, i.e. naltrexone hydrochloride. Naltrexone and its active metabolite 6β-naltrexol are competitive antagonists at μ- and κ-opioid receptors, and to a lesser extent at δ-opioid receptors. The plasma halflife of naltrexone is about 4 hours, for 6-β-naltrexol 13 hours.
One important step in the synthesis of naltrexone is a demethylation step for removing the methyl group from the nitrogen atom of the alkaloid molecule, as it is present for example in oxycodone and oxymorphone.
One way to achieve this demethylation step is disclosed in GB-1,124,441. Therein, a process of removing a methyl group from a tertiary nitrogen atom of an alkaloid molecule is disclosed, comprising reacting the methylated compound with a lower alkyl azodicarboxylate, wherein the lower alkyl group has from 1-8 carbon atoms, for a time sufficient to bring about a reaction and to provide a demethylated alkaloid derivative. However, the patent specification is only related to alkaloid compounds, which do not carry an OH-group at C14 of the alkaloid backbone. For example, GB-1,124,441 is silent on how to perform a demethylation on oxycodone and oxymorphone, respectively.
The paper titled “Reactions of Azodicarboxylic Esters with Amines” of S. HOSZTAFI, Österreichische Apotheker-Verlagsges.m.b.H., Wien, 1987, describes the reactions of aliphatic and aromatic amines and alkaloids with azodicarboxylic esters.
Further documents are available which are related to the demethylation of alkaloids, for example, EP 0 295 783 discloses diethyl azodicarboxylate for demethylating alkaloids. On page 8 of the description, it is disclosed to use diethyl azodicarboxylate in acetonitrile for the N-dealkylation of a compound of the indicated formula. However, in the course of this reaction, a bridging group is introduced between 2N-atoms as it can be seen from formula (VII). Furthermore, EP 0 295 783 is not related to alkaloids having a 14-OH group.
The like, GB-1,179,479 discloses removing a methyl group by treatment with a di-lower alkyl azodicarboxylate, wherein the lower alkyl group has from 1-4 carbon atoms, followed by treatment of delute mineral acid. This reaction, however, is disclosed in the context of general formula (I) as indicated in GB-1,179,479, which does not possess a 14-OH group.
In the prior art, there are also chemical processes disclosed, which allow the demethylation of alkaloid compounds which are carrying an OH-group at C14. In these methods, however, the OH group is being protected in order to avoid a reaction with the usual demethylation agents.
For example, oxycodone may be reacted with Ac2O in order to achieve acetyloxycodone (having a protected OH-group) which is further converted by well-known demethylating agents (BrCN/H2SO4) to noroxycodone (which is demethylated).
The same step can be performed in the synthesis of noroxymorphone, starting from oxymorphone. There, oxymorphone is reacted with Ac2O resulting in a protected OH-group at C14 (oxymorphone diacetate). After reacting with BrCN and H2SO4, noroxymorphone is yielded. The conversion by means of BrCN is described in IIJIMA et al., “Studies in the (+)-Morphinan Series. 5. Synthesis and Biological Properties of (+)-Naloxone”, Journal of Medicinal Chemistry, 1978, Vol. 21, No. 4.
However, there is not any disclosure in the art for a process for demethylating alkaloid derivatives which are carrying the 14-OH group, which 14-OH group is not being protected during the reaction. This might be due to the fact that conventional demethylating agents, such as BrCN, do not function under these circumstances as proper demethylating agents: The free OH group at C14 is too close to the cyanide and can perform a 5-exo-dig-cyclization which would not allow a proper performance of the reaction. See in this connection, CURRIE, A. C.; NEWBOLD, G. T. et al., Roy. Coll. Sci., Technol., Glasgow, UK, Journal of the Chemical Society, abstracts (1961), 4693-4700.
Further, see also GB-975,601, published in 1964. Here, it is disclosed that 14-acetoxy-N-cyanonorcodeine acetate may be produced by reacting 14-acetoxycodeine acetate with cyanogen bromide and may be converted to 14-hydroxynorcodeine by means of lithium aluminium hydride.